CN114549035A - Construction method of financial user accurate customer acquisition label based on telecommunication big data - Google Patents

Construction method of financial user accurate customer acquisition label based on telecommunication big data Download PDF

Info

Publication number
CN114549035A
CN114549035A CN202111633814.5A CN202111633814A CN114549035A CN 114549035 A CN114549035 A CN 114549035A CN 202111633814 A CN202111633814 A CN 202111633814A CN 114549035 A CN114549035 A CN 114549035A
Authority
CN
China
Prior art keywords
user
data
financial
model
big data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111633814.5A
Other languages
Chinese (zh)
Inventor
宋玉峰
李梦凡
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianyi Electronic Commerce Co Ltd
Original Assignee
Tianyi Electronic Commerce Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianyi Electronic Commerce Co Ltd filed Critical Tianyi Electronic Commerce Co Ltd
Priority to CN202111633814.5A priority Critical patent/CN114549035A/en
Publication of CN114549035A publication Critical patent/CN114549035A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0201—Market modelling; Market analysis; Collecting market data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/21—Design, administration or maintenance of databases
    • G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

The invention discloses a construction method of a financial user accurate visitor-obtaining label based on telecommunication big data, which mainly comprises the following steps: the method comprises three modules of data acquisition and cleaning, mobile APP selection and empowerment and index calculation, and more effective user data are acquired based on telecommunication big data; based on actual data of old users, the model and the index construction process are guided, and the empowerment and calculation method is more scientific; in the data cleaning process, a machine learning mode is used, and the feature data of the new user is clustered according to the characteristics of the data generated by the old user. The invention describes the financial attribute portrait of the user, such as the investable assets, investment experience, liveness and the like of the user; a model for evaluating the financial attributes of the user is used, behavior big data such as financial bank APP browsing and the like of the user are tracked, and then the cognition of the risk habits of the user is realized.

Description

Construction method of financial user accurate customer acquisition label based on telecommunication big data
Technical Field
The invention relates to the field of user portrait exploration, in particular to a construction method of a financial user accurate customer acquisition tag based on telecommunication big data.
Background
With the continuous abundance of financial products, the related theories of financial user operation and life cycle management are continuously mature, and a mature experience map and user label construction method is formed in the field of financial product target customer group delineation and marketing. However, in the construction methods of the models or the tags, the problems of 'inaccurate identification of the financial attributes of the user', 'unreasonable data selection and cleaning methods in the process of constructing the tags' and the like exist, so that how to find the data for accurately positioning the financial attributes of the user and scientifically clean the data to construct the tags becomes a research hotspot.
The user portrait and the user label constructed on the basis of the user portrait are used as an effective tool for describing the core product appeal and the effective purchasing power of a target user, the user portrait and the user label are widely applied in various fields, for example, the application in the E-commerce field, enterprises abstract behavior information, transaction information and personal basic information of the user into an atomic label, and the user image is materialized through logic combination and actual product characteristic fitting, so that targeted product service and marketing transformation are provided for the user. However, in the process of making a user image into a user atom label, the problems that sensitive data and objective data are limited in collection, user core preference cannot be described, user image construction distortion cannot be described and the like exist at present, and the problems are more common in scenes such as financial services and the like, so that the problems are inevitably caused that when service information is pushed, the user is not required by a user, and user experience is poor.
In conclusion, the invention designs a financial user accurate customer acquisition label construction method based on telecommunication big data, which takes a dynamic financial old user group as a sample, constructs a label model according to the information of the user browsing and using APP categories, the redemption behavior characteristics and the like, and accurately defines the potential user group of corresponding products.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a construction method of a financial user accurate customer-obtaining label based on telecommunication big data, which locates indexes of characteristics of 'investable products', 'investment experience' and the like of a user, establishes a label model according to information of browsing and using APP categories, redemption behavior characteristics and the like of the user, and accurately defines a potential user group of corresponding products.
In order to achieve the purpose, the invention is realized by the following technical scheme: a construction method of a financial user accurate customer acquisition label based on telecommunication big data comprises the following steps:
1. classifying the apps which are commonly used by each large platform and are related to financial management, and endowing the apps with rights according to the downloading amount, the using frequency, the using duration and the using condition of a platform financial user;
2. finding the most relevant characteristic values of the platform financing user through the model, such as the highest position holding amount, the maximum purchase applying amount, the risk level and the like;
3. performing data cleaning by using a common method, wherein default values are given to missing values, and abnormal values are eliminated;
4. establishing a table of the sorted data;
5. modeling by using a spark framework, scoring the users by mainly applying a logistic regression model, evaluating the fitting effect of the model by using an ROC (rock and rockwell), and evaluating whether feasibility exists or not by combining with the actual condition of the users;
6. the assets, investment experience and liveness evaluation scores which can be thrown by the user are output to a drop table, and a label is established, so that the label is convenient for service personnel to use;
7. and in the later stage, the continuously updated financing users are combined, more related features and apps are added, the model is continuously iterated, and the accuracy is improved.
The invention has the following beneficial effects:
1. stock user image analysis and machine learning are applied, a new index construction method and an APP selection scheme are set, and more stable output and more accurate data results are achieved;
2. and setting a grading index of the APP use habit, packaging and reading the APP use data of the user in the telecommunication big data on the premise of not acquiring the detailed data of the user, and further taking the user behavior data into consideration as a standard for measuring the property investable by the user and the like.
Drawings
The invention is described in detail below with reference to the drawings and the detailed description;
FIG. 1 is a flow chart of index construction according to the present invention
FIG. 2 is an APP behavior data read dependency table structure of the present invention;
FIG. 3 is a flow chart of the offline modeling of the present invention;
FIG. 4 is a diagram illustrating APP selection according to the present invention;
FIG. 5 is a representative illustration of the entitlement of the present invention;
FIG. 6 is a schematic diagram of an output user surmountable asset score of the present invention;
fig. 7 is a schematic diagram of the output user investment experience score of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
Referring to fig. 1 to 3, the following technical solutions are adopted in the present embodiment: a construction method of a financial user accurate customer acquisition label based on telecommunication big data comprises the following steps:
1. classifying the apps which are commonly used by each large platform and are related to financial management, and endowing the apps with rights according to the downloading amount, the using frequency, the using duration and the using condition of a platform financial user;
2. finding the most relevant characteristic values of the platform financing user through the model, such as the highest position holding amount, the maximum purchase applying amount, the risk level and the like;
3. performing data cleaning by using a common method, wherein default values are given to missing values, and abnormal values are eliminated;
4. establishing a table of the sorted data;
5. modeling by using a spark framework, scoring the users by mainly applying a logistic regression model, evaluating the fitting effect of the model by using an ROC (rock and rockwell), and evaluating whether feasibility exists or not by combining with the actual condition of the users;
6. the assets, investment experience and liveness evaluation scores which can be thrown by the user are output to a drop table, and a label is established, so that the label is convenient for service personnel to use;
7. and in the later stage, the continuously updated financing users are combined, more related features and apps are added, the model is continuously iterated, and the accuracy is improved.
The specific implementation mode is based on telecommunication big data, and more effective user data is obtained; based on actual data of old users, the model and the index construction process are guided, and the weighting and calculating method is more scientific; in the data cleaning process, a machine learning mode is used, and the feature data of the new user is clustered according to the characteristics of the data generated by the old user.
The specific implementation mode describes the financial attribute portrait of the user, such as assets investment, investment experience, liveness and the like of the user; a model for evaluating the financial attributes of the user is used, behavior big data such as financial bank APP browsing and the like of the user are tracked, and then the cognition of the risk habits of the user is realized.
The specific implementation mode can be applied to identification of target customer classification, customer risk grade, deliverable assets and the like of financial products; and outputting the marketing strategy of the financial products.
Example 1: the method for constructing the accurate customer acquisition tag of the financial user based on the telecommunication big data comprises the following steps:
first, the index construction is performed. With respect to investable asset metrics, financing users are differentiated according to the amount of assets taken, and non-financing users are trained by the data needed to use financing. Abstracting various atomic labels such as annual consumption, credit card consumption, telephone fee package amount and the like to represent the investable yield of non-financing users; risk preference financing users are classified according to financing risk evaluation results, non-financing users are trained through risk grade data of the financing users, financing provides a data source with a good risk preference, and big data experts evaluate and train and abstract various atom labels such as 'financing APP use frequency and teaching APP use frequency';
secondly, APP selection and offline modeling are carried out, and user income, payment, credit card amount and telephone fee package (the highest 100 points and the lowest 0 point) are obtained through simulation according to the following formulas; whether a vehicle exists or not, whether the vehicle is a family user (with score of 100 or not, without score of 0) and the like are evaluated, and the following data are calculated according to the highest score of 100 and the lowest score of 0:
score＝100*log(Value)/(log(Max))
in the final calculation model, the non-financing user investable yield score is 0.6, the average payment amount score of the last half-year month +0.4, the average payment income score of the last half-year month, and the financing user investable yield score is 0.5, the user historical highest position holding score +0.5, and the historical maximum purchase applying score.
Regarding the equal division of investment experience, the scores of the months of the use of the investment financing APP are assigned by A-E according to the calculation scores (the highest 100 points and the lowest 0 point) of the total months of use of the APP at each stage, and the assignment table is shown in FIG. 5:
score＝100*log(Value)/(log(Max))
the investment financing APP uses the monthly score as A-level weight A-level score + B-level weight B-level score + C-level weight C-level score + D-level weight D-level score + E-level weight E-level score.
According to the modeling process, the output user investable asset score (fig. 6) and investment experience score (fig. 7) are distributed as follows, and the manager financial user score is consistent with the distribution of the actual characterization data (taken positions, wind measurement questionnaires and the like).
The APP selection (taking the investment experience label as an example) and weight implementation of this embodiment are described as follows:
1. selecting the principle: taking user investment experience indexes as an example, selecting the APP frequently browsed by the user as a main target according to the inventory user as an example, and expanding and selecting the APP of the same type
2. The empowerment principle is as follows: sequencing according to browsing indexes (the browsing indexes refer to the use frequency of the APP by the old user; the indexes are high and the empowerment is high) corresponding to various APPs and user scale magnitude (the magnitude is large and the empowerment is low), and assigning weight scores from high to low according to the sequencing
3. And (3) under the condition that the old user has less data, performing model training by using machine learning, and simulating to obtain partial data results, such as comprehensive consideration on daily telephone charge package limit, location information and the like of the user in the process that the user can put assets, and fitting with the assets of the same type of old users.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (2)

1. A construction method of a financial user accurate customer acquisition label based on telecommunication big data is characterized by comprising the following steps:
(1) classifying the apps which are commonly used by each large platform and are related to financial management, and endowing the apps with rights according to the downloading amount, the using frequency, the using duration and the using condition of a platform financial user;
(2) finding the most relevant characteristic value of the platform financing user through the model;
(3) performing data cleaning by using a common method, wherein default values are given to missing values, and abnormal values are eliminated;
(4) establishing a table of the sorted data;
(5) modeling by using a spark framework, scoring the users by mainly applying a logistic regression model, evaluating the fitting effect of the model by using an ROC (rock and rockwell model), and evaluating whether feasibility exists or not by combining with the actual situation of the users;
(6) the assets, investment experience and liveness evaluation scores which can be thrown by the user are output to a drop table, and a label is established, so that the use of business personnel is facilitated;
(7) and in the later stage, the financial management users which are continuously updated are combined, more related features and apps are added, and the model is continuously iterated, so that the accuracy is improved.
2. The method for constructing the financial user precision customer label based on the telecom big data as claimed in claim 1, wherein the characteristic values in step (2) are highest position holding amount, maximum purchase requisition amount, risk level.
CN202111633814.5A 2021-12-28 2021-12-28 Construction method of financial user accurate customer acquisition label based on telecommunication big data Pending CN114549035A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111633814.5A CN114549035A (en) 2021-12-28 2021-12-28 Construction method of financial user accurate customer acquisition label based on telecommunication big data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111633814.5A CN114549035A (en) 2021-12-28 2021-12-28 Construction method of financial user accurate customer acquisition label based on telecommunication big data

Publications (1)

Publication Number Publication Date
CN114549035A true CN114549035A (en) 2022-05-27

Family

ID=81669254

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111633814.5A Pending CN114549035A (en) 2021-12-28 2021-12-28 Construction method of financial user accurate customer acquisition label based on telecommunication big data

Country Status (1)

Country Link
CN (1) CN114549035A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115760200A (en) * 2023-01-06 2023-03-07 万链指数(青岛)信息科技有限公司 User portrait construction method based on financial transaction data

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115760200A (en) * 2023-01-06 2023-03-07 万链指数(青岛)信息科技有限公司 User portrait construction method based on financial transaction data
CN115760200B (en) * 2023-01-06 2023-07-04 万链指数(青岛)信息科技有限公司 User portrait construction method based on financial transaction data

Similar Documents

Publication Publication Date Title
CN109255506B (en) Internet financial user loan overdue prediction method based on big data
CN106651424B (en) Power user portrait establishing and analyzing method based on big data technology
Chien et al. Manufacturing intelligence for semiconductor demand forecast based on technology diffusion and product life cycle
CN109409677A (en) Enterprise Credit Risk Evaluation method, apparatus, equipment and storage medium
Cheng et al. Customer lifetime value prediction by a Markov chain based data mining model: Application to an auto repair and maintenance company in Taiwan
US20030097292A1 (en) System and method for stability analysis of profitability of target markets for goods or services
CN108388974A (en) Top-tier customer Optimum Identification Method and device based on random forest and decision tree
CN109711424B (en) Behavior rule acquisition method, device and equipment based on decision tree
CN109636482B (en) Data processing method and system based on similarity model
CN112837087A (en) User portrait construction method oriented to consultation service system
CN107133315A (en) A kind of smart media based on semantic analysis recommends method
CN113469730A (en) Customer repurchase prediction method and device based on RF-LightGBM fusion model under non-contract scene
CN108154311A (en) Top-tier customer recognition methods and device based on random forest and decision tree
CN111882420A (en) Generation method of response rate, marketing method, model training method and device
CN112102006A (en) Target customer acquisition method, target customer search method and target customer search device based on big data analysis
Sun et al. Using improved RFM model to classify consumer in big data environment
CN114549035A (en) Construction method of financial user accurate customer acquisition label based on telecommunication big data
Byrne The digital economy and productivity
CN112330426A (en) Product recommendation method, device and storage medium
CN116663909A (en) Provider risk identification data processing method and device
CN116579799A (en) Processing and modulating method and system applied to pet puffed food preparation
CN116187808A (en) Electric power package recommendation method based on virtual power plant user-package label portrait
CN115293867A (en) Financial reimbursement user portrait optimization method, device, equipment and storage medium
CN115048503A (en) User preference label design method based on content analysis
Fang et al. Data mining technology and its Application In CRM of Commercial Banks

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication